Sandip Aine

Shohin Mukherjee, Sandip Aine, Maxim Likhachev

Lazy search algorithms have been developed to efficiently solve planning problems in domains where the computational effort is dominated by the cost of edge evaluation. The existing algorithms operate by intelligently balancing computational effort between searching the graph and evaluating edges. However, they are designed to run as a single process and do not leverage the multithreading capability of modern processors. In this work, we propose a massively parallelized, bounded suboptimal, lazy search algorithm (MPLP) that harnesses modern multi-core processors. In MPLP, searching of the graph and edge evaluations are performed completely asynchronously in parallel, leading to a drastic improvement in planning time. We validate the proposed algorithm in two different planning domains: 1) motion planning for 3D humanoid navigation and 2) task and motion planning for a robotic assembly task. We show that MPLP outperforms the state-of-the-art lazy search as well as parallel search algorithms. The open-source code for MPLP is available here: https://github.com/shohinm/parallel_search

Ramkumar Natarajan, Muhammad Suhail Saleem, William Xiao et al.

Designing good heuristic functions for graph search requires adequate domain knowledge. It is often easy to design heuristics that perform well and correlate with the underlying true cost-to-go values in certain parts of the search space but these may not be admissible throughout the domain thereby affecting the optimality guarantees of the search. Bounded suboptimal search using several such partially good but inadmissible heuristics was developed in Multi-Heuristic A* (MHA*). Although MHA* leverages multiple inadmissible heuristics to potentially generate a faster suboptimal solution, the original version does not improve the solution over time. It is a one shot algorithm that requires careful setting of inflation factors to obtain a desired one time solution. In this work, we tackle this issue by extending MHA* to an anytime version that finds a feasible suboptimal solution quickly and continually improves it until time runs out. Our work is inspired from the Anytime Repairing A* (ARA*) algorithm. We prove that our precise adaptation of ARA* concepts in the MHA* framework preserves the original suboptimal and completeness guarantees and enhances MHA* to perform in an anytime fashion. Furthermore, we report the performance of A-MHA* in 3-D path planning domain and sliding tiles puzzle and compare against MHA* and other anytime algorithms.

Guru Koushik Senthil Kumar, Sandip Aine, Maxim Likhachev

Paths planned over grids can often be suboptimal in an Euclidean space and contain a large number of unnecessary turns. Consequently, researchers have looked into post-processing techniques to improve the paths after they are planned. In this paper, we propose a novel post-processing technique, called Homotopic Visibility Graph Planning (HVG) which differentiates itself from existing post-processing methods in that it is guaranteed to shorten the path such that it is at least as short as the provably shortest path that lies within the same topological class as the initially computed path. We propose the algorithm, provide proofs and compare it experimentally against other post-processing methods and any-angle planning algorithms.

Manash Pratim Das, Anirudh Vemula, Mayank Pathak et al.

Consider a truck filled with boxes of varying size and unknown mass and an industrial robot with end-effectors that can unload multiple boxes from any reachable location. In this work, we investigate how would the robot with the help of a simulator, learn to maximize the number of boxes unloaded by each action. Most high-fidelity robotic simulators like ours are time-consuming. Therefore, we investigate the above learning problem with a focus on minimizing the number of simulation runs required. The optimal decision-making problem under this setting can be formulated as a multi-class classification problem. However, to obtain the outcome of any action requires us to run the time-consuming simulator, thereby restricting the amount of training data that can be collected. Thus, we need a data-efficient approach to learn the classifier and generalize it with a minimal amount of data. A high-fidelity physics-based simulator is common in general for complex manipulation tasks involving multi-body interactions. To this end, we train an optimal decision tree as the classifier, and for each branch of the decision tree, we reason about the confidence in the decision using a Probably Approximately Correct (PAC) framework to determine whether more simulator data will help reach a certain confidence level. This provides us with a mechanism to evaluate when simulation can be avoided for certain decisions, and when simulation will improve the decision making. For the truck unloading problem, our experiments show that a significant reduction in simulator runs can be achieved using the proposed method as compared to naively running the simulator to collect data to train equally performing decision trees.